Event Mapper

Traditional applications, which work with data at rest, and Event Processing Applications, which work with data in motion, may need to share data via an Event Streaming Platform. These applications need a common mechanism to convert data from Events to domain objects and vice versa.

Problem

How can I move data between domain objects in an application’s internal data model and Events in an Event Streaming Platform, while keeping the two independent of each other?

Solution

An Event Mapper provides independence between the traditional application and the Event Streaming Platform, so that neither is aware of the other (and ideally, neither is even aware of the event mapper itself).

Create an Event Mapper, or use an existing one, to map the domain model (or the application's internal data model) to the data formats accepted by the Event Streaming Platform, and vice versa. The Event Mapper reads the domain model and converts it into outgoing Events, which are sent to the Event Streaming Platform. Conversely, an Event Mapper can be used to create or update domain objects based on incoming Events.

Implementation

In this example, we use the Java producer client of Apache Kafka® to implement an Event Mapper that constructs an Event (PublicationEvent) from the domain model object (Publication) before the event is written to an Event Stream (called a "topic" in Kafka).

private final IMapper domainToEventMapper = mapperFactory.buildMapper(Publication.class);
private final Producer<String, PublicationEvent> producer = ...

public void newPublication(String author, String title) {
  Publication newPub = new Publication(author, title);
  producer.send(author /* event key */, domainToEventMapper.map(newPub));

We can implement the reverse operation in a second Event Mapper that converts PublicationEvent Events back into domain object updates:

private final IMapper eventToDomainMapper = mapperFactory.buildMapper(Publication.class);
private final Consumer<String, PublicationEvent> consumer = ...

public void updatePublication(PublicationEvent pubEvent) {
  Publication newPub = eventToDomainMapper.map(pubEvent);
  domainStore.update(newPub);

Considerations

The Event Mapper may optionally validate the schema of the converted objects. For details, see the Schema Validator pattern.

References

• This pattern is derived from Messaging Mapper in Enterprise Integration Patterns, by Gregor Hohpe and Bobby Woolf.
• See also the Event Serializer and Event Deserializer patterns.